Automatic detecting actomyosin complex biological particles in cryo-EM images

This paper describes how actomyosin complex particles are detected automatically. We propose a new approach that combines both gray level co-occurrence matrix to extract texture features and the SVM classifier to detect actomyosin complex particles automatically. Experimental results show that detection rate achieves 94.81%, the false positive rate is 2.79%, and the false negative rate is 5.46%.